LLMs and its output are bounded by Rices theorem. This is not going to ensure correctness it’s just going to validate that the model can produce an undecidable result.
LLMs and its output are bounded by Rices theorem. This is not going to ensure correctness it’s just going to validate that the model can produce an undecidable result.
Errr, checking correctness of proofs is decidable.